Intra-operative surgical instrument usage detection on a multi-sensor table

Bernhard Glaser,Stefan Dänzer,Thomas Neumuth

DOI: https://doi.org/10.1007/s11548-014-1066-0

2014-05-15

International Journal of Computer Assisted Radiology and Surgery

Abstract:Purpose Automatic detection, classification and recording of operating room (OR) activities in real time during a surgical procedure requires a wide range of sensors to gather information on the activities of the surgeon and staff, the patient, and the OR equipment. The surgical instrument currently being used is an important parameter used to assess the instantaneous operating room status. An automated system was developed that detects unmodified surgical instruments in real time using a sensor-based table.Methods A multi-sensor operating room table was designed featuring a 2D camera, digital scale, and infrared camera. Software was developed to detect and record the sequence of changes on the table during a surgical procedure. The detection rates were evaluated under laboratory conditions by recording the observed instrument usage in 27 functional endoscopic sinus surgeries (FESS).Results  The detection rate achieved using video-based detection alone was 84.9 %. The total detection rate achieved with the combined approach using both video- and weight-based information was 90.3 %.Conclusion A multi-sensor table-based automated instrument tracking system was developed that provides a foundation for the intra-operative detection of surgical instruments without modifying the instruments in the surgical tray. This system was tested and found to satisfy clinical FESS requirements with a reasonable accuracy. This system may also be useful to improve patient safety, for example to prevent instruments being left in the patient.

engineering, biomedical,radiology, nuclear medicine & medical imaging,surgery